Hybrid ultramicroporous materials (HUMs) are porous coordination networks composed of combinations of organic and inorganic linker ligands with a pore diameter of <7 Å. Despite their benchmark gas sorption selectivity for several industrially relevant gas separations and their inherent modularity, the structural and compositional diversity of HUMs remains underexplored. In this contribution, we report a family of six HUMs (SIFSIX-22-Zn, TIFSIX-6-Zn, SNFSIX-2-Zn, GEFSIX-4-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn) based on Zn metal centers and the tetratopic N-donor organic ligand tetra(4-pyridyl)benzene (tepb). The incorporation of fluorinated inorganic pillars (SiF6 2-, TiF6 2-, SnF6 2-, GeF6 2-, ZrF6 2-, and TaF7 2-, respectively) resulted in (4,6)-connected fsc topology as verified using single-crystal X-ray diffraction. Pure-component gas sorption studies with N2, CO2, C2H2, C2H4, and C2H6 revealed that the large voids and narrow pore windows common to all six HUMs can be leveraged to afford high C2H2 uptakes while retaining high ideal adsorbed solution theory (IAST) selectivities for industrially relevant gas mixtures: >10 for 1:99 C2H2/C2H4 and >5 for 1:1 C2H2/CO2. The approach taken, systematic variation of pillars with retention of structure, enables differences in selectivity to be attributed directly to the choice of the inorganic pillar. This study introduces fsc topology HUMs as a modular platform that is amenable to fine-tuning of structure and properties.
Hybrid ultramicroporous materials (HUMs) are porous coordination networks composed of combinations of organic and inorganic linker ligands with a pore diameter of <7 Å. Despite their benchmark gas sorption selectivity for several industrially relevant gas separations and their inherent modularity, the structural and compositional diversity of HUMs remains underexplored. In this contribution, we report a family of six HUMs (SIFSIX-22-Zn, TIFSIX-6-Zn, SNFSIX-2-Zn, GEFSIX-4-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn) based on Zn metal centers and the tetratopic N-donor organic ligand tetra(4-pyridyl)benzene (tepb). The incorporation of fluorinated inorganic pillars (SiF6 2-, TiF6 2-, SnF6 2-, GeF6 2-, ZrF6 2-, and TaF7 2-, respectively) resulted in (4,6)-connected fsc topology as verified using single-crystal X-ray diffraction. Pure-component gas sorption studies with N2, CO2, C2H2, C2H4, and C2H6 revealed that the large voids and narrow pore windows common to all six HUMs can be leveraged to afford high C2H2 uptakes while retaining high ideal adsorbed solution theory (IAST) selectivities for industrially relevant gas mixtures: >10 for 1:99 C2H2/C2H4 and >5 for 1:1 C2H2/CO2. The approach taken, systematic variation of pillars with retention of structure, enables differences in selectivity to be attributed directly to the choice of the inorganic pillar. This study introduces fsc topology HUMs as a modular platform that is amenable to fine-tuning of structure and properties.
[Zn(4,4′-bipyridine)2(SiF6)] (SIFSIX-1-Zn), reported in 1995,[1] is the prototypal
hybrid coordination network
(HCN) and is the parent of what is today a broad and growing platform
of porous materials with a diverse range of topologies, pore sizes,
pore chemistries, and properties.[2] That
both the pore size and pore chemistry of HCNs are amenable to fine-tuning
through crystal engineering approaches is a consequence of their inherently
modular nature, which comes from their typical compositions: a divalent
metal ion node, neutral N-donor linker ligand, and
inorganic anion pillar.[2] The resulting
hybrid coordination networks are thereby composed of geometrically
simple components.[3] The ready availability
of N-donor linker ligands of varying lengths (e.g., pyrazine, 2.8 Å; 4,4′-bipyridine, 7.1
Å; N,N′-di(4-pyridyl)-1,4,5,8-naphthalenediimide, 15.4 Å) and inorganic
dianions that can serve as linkers that offer strong electrostatics
(e.g., SiF62–, GeF62–, AlF52–, NbOF52–, MoO42–, Cr2O72–) has been exploited to fine-tune properties
that are of relevance to several important industrial separation challenges.[4−11] In particular, HCNs with pore diameters <7 Å (hybrid ultramicroporous
materials, HUMs) exhibit selectivity values that are in some cases
orders of magnitude greater than previous benchmark porous materials: SIFSIX-18-Ni-β, NbOFFIVE-1-Ni, and TIFSIX-3-Ni for direct air capture of CO2; SIFSIX-14-Cu-i for removal of acetylene from ethylene; and CROFOUR-1-Ni for Xe/Kr separation.[12−22]Although crystal engineering has enabled systematic access
to platforms
of HUMs with exceptional properties, there is limited topological
diversity among the HUMs reported thus far. The majority of HCNs,
including HUMs, are pcu topology nets composed of octahedral
metal centers linked by two ditopic N-donor linker
ligands and a ditopic inorganic pillar anion. Among the non-pcu topology HCNs, only those that exhibit mmo topology have been studied systematically.[23] Prior to the recent reports of ZJU-280 (SIFSIX-22-Cu, [Cu(tepb)SiF6]) and TIFSIX-Cu-TPB (TIFSIX-6-Cu, [Cu(tepb)TiF6]), fsc-2-SIFSIX ([Cu3(4-(pyridin-4-yl)acrylic acid)4(SiF6)]), CPM-131 ([(TPyP-Fe)Zn(SiF6)]),
and its analogues ([(TPyP-M)Cu(NbOF5)], M = Zn, Fe, Ni)
were the only HCNs with (4,6)-connected fsc topology,
and their sorption properties were found to be driven by coordinatively
unsaturated metal centers (UMCs) rather than electrostatics and tight
binding sites.[24−28] In fsc-2-SIFSIX, a bifunctional organic linker ligand
allows for incorporation of both mononuclear Cu(II) and dinuclear
{Cu2} paddle-wheel building blocks into the final structure.[28]CPM-131 (and the related fsx net CPM-132) is constructed using a porphyrin-based
metalloligand, and despite the challenges of tuning a porphyrin-based
system, it exemplifies HCNs based on a polytopic (used herein to refer
to connectivity >2) ligand.[25,26] In 2016, Lusi et al.
reported a family of HCNs based on the polytopic linker Tripp (2,4,6-tris(4-pyridyl)pyridine), Tripp-Cu-MFSIX ([[Cu6(Tripp)8](MF6)3(MF6)3]).[29] These structures formed a partially bridged pto-type net but were not found to be permanently porous despite large
solvent-accessible voids. A notable feature of this platform is that
five distinct inorganic pillar dianions were incorporated into the
same structure. More recently, Wu et al. reported
an ith-d topology framework, SIFSIX-Cu-TPA ([Cu3(TPA)4(SiF6)3]),
using a tritopic linker ligand, TPA (tri(pyridin-4-yl)amine).[30] Our group recently reported the fsc frameworks SIFSIX-22-Zn and SOFOUR-1-Zn, both of which are based on the tepb linker.[31]The modularity of a porous coordination network (PCN)
can be expressed
in terms of how many components can be varied independently. HCNs
based on two-dimensional nets pillared by MFSIX to form pcu networks are highly modular, having three components that can be
varied independently (organic linker ligand, inorganic anionic pillar
ligand, and metal cation node).[2] Additionally,
there are cases in which interpenetration can also be controlled,
for example, in SIFSIX-2-Cu and SIFSIX-2-Cu-i.[4] The resulting drastic effect on the
pore size and pore chemistry that results from interpenetration (and
therefore properties) offers a fourth variable by which such a platform
can be modulated. This level of modularity (four components) is only
met or surpassed by platforms that combine mixed linkers and/or extra-framework
anions/cations. Most other well-known PCNs, such as those based on
oxo-clusters and carboxylate linkers, offer just one or two modular
components, limiting the scope of related materials that can be generated
and, therefore, the extent to which properties can be “fine-tuned”.A crystal engineering approach predominantly based on ditopic linker
ligands means that most HCNs exhibit nearly cylindrical one-dimensional
channels with a high density of tight binding sites to drive their
sorption properties. The design of HCNs based on polytopic ligands
offers the possibility of new types of channel architectures and new
structure–property relationships. In their recent work on ZJU-280 (SIFSIX-22-Cu), Qian and co-workers reported a HUM
that is composed of a tetratopic linker ligand in place of the more
commonly used ditopic linkers, presenting an opportunity for the development
of a new HUM platform for the study of structure–property relationships.[24] We recently reported SIFSIX-22-Zn and SOFOUR-1-Zn using the same tetrapyridyl linker
and SiF62– or SO42– pillars, respectively.[31] In the present
work, we report a crystal engineering study of fsc HUMs
involving substitution of the inorganic pillars in [Zn(tepb)SiF6] (SIFSIX-22-Zn) to afford an additional five
members of this platform: [Zn(tepb)TiF6] (TIFSIX-6-Zn), [Zn(tepb)SnF6] (SNFSIX-2-Zn), [Zn(tepb)GeF6] (GEFSIX-4-Zn), [Zn(tepb)ZrF6] (ZRFSIX-3-Zn), and [Zn(tepb)TaF7] (TAFSEVEN-1-Zn).
Experimental Section
Materials and Methods
All reagents and solvents were
used as received from vendors. 1H NMR spectroscopy was
performed using a JEOL ECX400 spectrometer operating at 400 MHz. Thermal
gravimetric analysis (TGA) was performed using a TA Q50 analyzer with
a ramp rate of 10.00 °C/min from 25 to 500 °C and nitrogen
gas flow of 40 mL/min. Powder X-ray diffraction (PXRD) diffractograms
were recorded using a PANalytical X’Pert operated at 40 kV
and 40 mA and CuKα radiation (λα = 1.540598 Å) was used for diffraction experiments. Incident
beam optics included the Fixed Divergence slit with antiscatter slit
PreFIX module, with a 1/8° divergence slit and a 1/4° antiscatter
slit, as well as a 10 mm fixed incident beam mask and a Ni-β
filter. Data were collected from 5°–40° (2θ)
with a step-size of 0.0131303° and a scan time of 30 s per step.
Synthesis of 1,2,4,5-Tetra(4-pyridyl)benzene, tepb
1,2,4,5-Tetra(4-pyridyl)benzene (tepb) was synthesized
according to the procedure reported by Chang and Wang.[32] Fe(NO3)3·9H2O (0.161 g, 0.4 mmol), H3PO4 (0.6 mL, 9 mmol),
1,3-bis(4-pyridyl)propane (1.268 g, 6.4 mmol), oxalic acid dihydrate
(0.151 g, 1.2 mmol), and water (2 mL) were combined in a Teflon-lined
pressure vessel and heated at 180 °C for 48 h. Needle crystals
of tepb were manually removed, washed with cold methanol, and dried
(yield: ca. 50%). 1H NMR (400 MHz, DMSO-d6): δ = 8.48 (8H, d), 7.61 (2H, s), 7.24 (8H, d).
Synthesis of [Zn(tepb)SiF6], SIFSIX-22-Zn
A solution of ZnSiF6·6H2O (1.0
mg, 0.003 mmol) in 0.4 mL of methanol was put in a narrow glass tube.
Methanol (0.2 mL) was carefully layered over this solution to act
as a buffer layer before a solution of tepb (1.2 mg,
0.003 mmol) in 0.4 mL methanol was layered over the buffer layer.
The tube was left undisturbed for 5 days, at which point colorless
block crystals of [Zn(tepb)SiF6]·xMeOH (as-synthesized SIFSIX-22-Zn) were obtained. A
larger quantity of SIFSIX-22-Zn was prepared as follows:
ZnSiF6·6H2O (41.2 mg, 0.13 mmol) was added
to a solution of tepb (77.2 mg, 0.20 mmol) in 16 mL methanol
and stirred at room temperature overnight. [Zn(tepb)SiF6]·xMeOH was obtained as a white microcrystalline
powder which was isolated by filtration before being washed with methanol
and air dried. Yield: 43.0 mg, 56%. CHN analysis calculated for C26H26F6N4O4SiZn
(including four interstitial water molecules): C 46.89%, H 3.94%,
N 8.41%; experimental: C 46.72%, H 3.12%, N 8.27%.
Synthesis of [Zn(tepb)TiF6], TIFSIX-6-Zn
A solution of Zn(NO3)2·6H2O (0.89 mg, 0.003 mmol) and (NH4)2TiF6 (0.59 mg, 0.003 mmol) in 0.2 mL of water was placed in a
narrow glass tube. 1:1 methanol/water (0.4 mL) was carefully layered
over this solution to act as a buffer layer before a solution of tepb (1.2 mg, 0.003 mmol) in 0.4 mL methanol was layered over
the buffer layer, and the tube was left undisturbed for 3 days, at
which point colorless block crystals of [Zn(tepb)TiF6]·xMeOH (as-synthesized TIFSIX-6-Zn) were obtained.
A bulk sample of TIFSIX-6-Zn was prepared as follows:
Zn(NO3)2·6H2O (77.3 mg, 0.26
mmol) and (NH4)2TiF6 (51.5 mg, 0.26
mmol) in 1.0 mL of water was added to a solution of tepb (154.4 mg, 0.4 mmol) in 30 mL methanol and stirred at room temperature
overnight. [Zn(tepb)TiF6]·xMeOH was
obtained as a white microcrystalline powder, isolated by filtration,
washed with methanol, and air-dried. Yield: 96 mg, 60%. CHN analysis
calculated for C26H28F6N4O5TiZn (including five interstitial water molecules):
C 44.37%, H 4.01%, N 7.96%; experimental: C 44.37%, H 3.41%, N 8.11%.
Synthesis of [Zn(tepb)GeF6], GEFSIX-4-Zn
A solution of Zn(NO3)2·6H2O (0.89 mg, 0.003 mmol) and (NH4)2GeF6 (0.66 mg, 0.003 mmol) in 0.2 mL of water was placed in a
narrow glass tube. 1:1 methanol/water (0.4 mL) was carefully layered
over this solution to act as a buffer layer. Finally, a solution of tepb (1.2 mg, 0.003 mmol) in 0.4 mL methanol was layered over
the buffer layer, and the tube was left undisturbed for 3 days. Colorless
block crystals of [Zn(tepb)GeF6]·xMeOH (as-synthesized GEFSIX-4-Zn) were thereby obtained.
A bulk sample of GEFSIX-4-Zn was prepared as follows:
Zn(NO3)2·6H2O (77.3 mg, 0.26
mmol) and (NH4)2GeF6 (57.9 mg, 0.26
mmol) in 1.0 mL of water was added to a solution of tepb (154.4 mg, 0.4 mmol) in 30 mL methanol and stirred at room temperature
overnight. [Zn(tepb)GeF6]·xMeOH was
obtained as a white microcrystalline powder, which was isolated by
filtration, washed with methanol, and air-dried. Yield: 108 mg, 65%.
CHN analysis calculated for C26H26F6N4O4GeZn (including four interstitial water
molecules): C 43.95%, H 3.69%, N 7.87%; experimental: C 43.95%, H
3.24%, N 7.89%.
Synthesis of [Zn(tepb)SnF6], SNFSIX-2-Zn
A solution of Zn(NO3)2·6H2O (0.89 mg, 0.003 mmol) and (NH4)2SnF6 (0.80 mg, 0.003 mmol) in 0.2 mL of water was placed in a
narrow glass tube. 1:1 methanol/water (0.4 mL) was carefully layered
over this solution to act as a buffer layer. Finally, a solution of tepb (1.2 mg, 0.003 mmol) in 0.4 mL methanol was layered over
the buffer layer, and the tube was left undisturbed for 3 days. Colorless
block crystals of [Zn(tepb)SnF6]·xMeOH (as-synthesized SNFSIX-2-Zn) were thereby obtained.
A bulk sample of SNFSIX-2-Zn was prepared as follows:
a solution Zn(NO3)2·6H2O (77.3
mg, 0.26 mmol) and (NH4)2SnF6 (69.4
mg, 0.26 mmol) in 1.0 mL of water was added to a solution of tepb (154.4 mg, 0.4 mmol) in 30 mL methanol and stirred at
room temperature overnight. [Zn(tepb)SnF6]·xMeOH was obtained as a white microcrystalline powder, which
was isolated by filtration, washed with methanol, and air-dried. Yield:
119 mg, 67%. CHN analysis calculated for C26H34F6N4O8SnZn (including eight interstitial
water molecules): C 37.69%, H 4.14%, N 6.76%; experimental: C 37.57%,
H 3.20%, N 6.75%.
Synthesis of [Zn(tepb)ZrF6], ZRFSIX-3-Zn
A solution of Zn(NO3)2·6H2O (0.89 mg, 0.003 mmol) and K2ZrF6 (0.85
mg, 0.003 mmol) in 0.2 mL of water was placed in a narrow glass tube.
1:1 methanol/water (0.4 mL) was carefully layered over this solution
to act as a buffer layer. Finally, a solution of tepb (1.2 mg, 0.003 mmol) in 0.4 mL methanol was layered over the buffer
layer, and the tube was left undisturbed for 3 days. Colorless block
crystals of [Zn(tepb)ZrF6]·xMeOH
(as-synthesized ZRFSIX-3-Zn) were thereby obtained. A
bulk sample of ZRFSIX-3-Zn was prepared as follows: A
solution Zn(NO3)2·6H2O (77.3
mg, 0.26 mmol) and K2ZrF6 (73.7 mg, 0.26 mmol)
in 10.0 mL of water was added to a solution of tepb (154.4
mg, 0.4 mmol) in 30 mL methanol and stirred at room temperature overnight.
[Zn(tepb)ZrF6]·xMeOH was obtained
as a white microcrystalline powder, which was isolated by filtration,
washed with methanol, and air-dried. Yield: 116 mg, 68%. CHN analysis
calculated for C26H28F6N4O5ZrZn (including five interstitial water molecules):
C 41.80%, H 3.78%, N 7.50%; experimental: C 41.92%, H 2.91%, N 7.82%.
Synthesis of [Zn(tepb)TaF7], TAFSEVEN-1-Zn
A solution of Zn(NO3)2·6H2O (0.89 mg, 0.003 mmol) and (NH4)2TaF7 (1.05 mg, 0.003 mmol) in 0.2 mL of water was placed in a
narrow glass tube. 1:1 methanol/water (0.4 mL) was carefully layered
over this solution to act as a buffer layer. Finally, a solution of tepb (1.2 mg, 0.003 mmol) in 0.4 mL methanol was layered over
the buffer layer, and the tube was left undisturbed for 3 days. Colorless
block crystals of [Zn(tepb)TaF7]·xMeOH (as-synthesized TAFSEVEN-1-Zn) were thereby obtained.
A bulk sample of TAFSEVEN-1-Zn was prepared as follows:
a solution Zn(NO3)2·6H2O (77.3
mg, 0.26 mmol) and (NH4)2TaF7 (91.0
mg, 0.26 mmol) in 1 mL of water was added to a solution of tepb (154.4 mg, 0.4 mmol) in 30 mL methanol and stirred at room temperature
overnight. [Zn(tepb)TaF7]·xMeOH was
obtained as a white microcrystalline powder, which was isolated by
filtration, washed with methanol, and air-dried. Yield: 113 mg, 57%.
CHN analysis calculated for C26H22F6N4O2TaZn (including two interstitial water
molecules): C 38.95%, H 2.77%, N 6.99%; experimental: C 38.81%, H
2.25%, N 7.05%.
X-ray Crystallography
Single-crystal X-ray crystallographic
data were collected at 298 K on a Bruker D8 Quest diffractometer equipped
with a CuKα microfocus source (λ = 1.5406 Å) and
Photon 100 detector. Temperature was controlled using a nitrogen flow
from Oxford Cryosystems. Data was indexed, integrated and scaled in
APEX3.[33] Absorption correction was performed
by the multi-scan method using SADABS.[34] Space group determination was performed simultaneously with structure
solution using the intrinsic phasing method (SHELXT),[35] and the solution was refined on F2 using SHELXL[36] nonlinear least
squares implemented in Olex2 v1.2.10.[37] All nonhydrogen atoms were refined anisotropically and hydrogen
atoms bonded to carbon atoms were added at calculated positions and
refined using a riding model. Disordered solvents were found in the
cavity of all structures. Some of this electron density could be modeled
as methanol molecules with partial occupancy; however, refinement
was unsatisfactory, and the solvent atomic displacement parameters
were unreasonable. The PLATON SQUEEZE[38] routine was performed to account for the electron density of unmodelled
solvents, resulting in a more satisfactory refinement. The crystal
structure CIF files have been deposited in the Cambridge Crystallographic
Data Centre (CCDC: 2151309–2151313).
Gas Sorption Measurements
For gas sorption experiments,
ultra-high purity gases were used as received from BOC Gases Ireland:
He (99.999%), CO2 (99.995%), C2H2 (98.5%), N2 (99.998%), C2H4 (99.92%),
and C2H6 (99.0%). Adsorption isotherm experiments
(up to 1 bar) for 195 K CO2 were performed on a Micromeritics
Tristar II 3030. A Micromeritics 3Flex surface area and pore size
analyzer 3500 was used for collecting the 273 and 298 K sorption isotherms
for all gases. The low temperature of 195 K was maintained using a
dry ice-acetone mixture. Bath temperatures of 273 and 298 K were precisely
controlled with a Julabo ME (v.2) recirculating control system containing
a mixture of ethylene glycol and water. Prior to experiments, SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn were activated on a Smart VacPrep using dynamic vacuum and heating
for 24 h at 333 K.
Results and Discussion
Structural Description
TIFSIX-6-Zn crystallized
in the centrosymmetric orthorhombic space group Pmma, SIFSIX-22-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, and TAFSEVEN-1-Zn crystallized in the centrosymmetric
orthorhombic space group Cmma while ZRFSIX-3-Zn crystallized in the centrosymmetric orthorhombic space group Pmmm. All six structures are comprised as expected: octahedral
Zn2+ ions coordinated to four pyridyl moieties of tepb ligands in their equatorial positions and bridging MF62– (M = Si(IV), Ti(IV), Ge(IV), Sn(IV),
and Zr(IV)) or TaF72– anions in their
axial positions. The Zn2+ ions and tepb ligands
formed two-dimensional layers pillared by the MF62– or TaF72– anions to generate three-dimensional
4,6-connected fsc topology networks (Figure ).
Figure 1
Representations of (a)
tetra(4-pyridyl)benzene (tepb) ligand and
Zn2+ metal center; (b) MFSIX and TAFSEVEN pillars; and (c) fsc network [Zn(tepb)MF6] viewed along the c (above) and a (below) crystallographic axes.
Representations of (a)
tetra(4-pyridyl)benzene (tepb) ligand and
Zn2+ metal center; (b) MFSIX and TAFSEVEN pillars; and (c) fsc network [Zn(tepb)MF6] viewed along the c (above) and a (below) crystallographic axes.The Zn–F distance in TAFSEVEN-1-Zn is 2.040(2)
Å, which lies within the lower quintile (2.061 Å) of the
mean distance of 2.108 Å (st. dev. = 0.059 Å) as per the
Cambridge Structural Database[39] (CSD, v.
5.41 + 3 updates, see the SI for search
parameters), while the Zn–F distances for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, and ZRFSIX-3-Zn lie in the second quintile (2.092 Å).
The TAFSEVEN-1-Zn Zn–N distance range of 2.176(3)
Å lies within the upper quintile (2.157 Å) of the mean distance
of 2.138 Å (st. dev. = 0.020 Å) as per the same search query.
Zn–N distances for GEFSIX-4-Zn and SNFSIX-2-Zn lie in the third quintile (2.145 Å) while the Zn–N distances
for SIFSIX-22-Zn, TIFSIX-6-Zn, and ZRFSIX-3-Zn lie in the fourth quintile (2.157 Å).The pyridine rings of the tepb ligand adopt a propeller-type
arrangement about the central benzene ring with rings para to each other being co-planar in TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn (Figure ). The torsion angles of the pyridine ring
about the Zn2+ ions are given in Figures S1–S4 and range from 52.3(2)° to 70.11(5)°.
The pyridyl rings are arranged in a propeller-like conformation about
the zinc(II) ion. The two-dimensional Zn2+-tepb layers are pillared by the MF62– and
TaF72– pillars such that the central
aromatic rings of the tepb ligands are eclipsed and coplanar
when viewed down the crystallographic a-axis for SIFSIX-22-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn and the
crystallographic b-axis for TIFSIX-6-Zn. In the case of TIFSIX-6-Zn and ZRFSIX-3-Zn, the fluorine atoms are modeled as disordered over two positions
with each position being eclipsed with the fluorine atoms above and
below the plane. For SIFSIX-22-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, and TAFSEVEN-1-Zn, the equatorial
fluorine atoms of the MF2– pillar are not eclipsed with GEFSIX-4-Zn having the
largest pillar rotation angle of 55.0(2)°. In the case of SNFSIX-2-Zn, the equatorial fluorine atoms of the SnF62– pillar are modeled as disordered over
two positions with each position exhibiting noticeably different pillar
rotation angles: 4.9(5)° and 23.0(5)°. In the two-dimensional
zinc(II)-tepb layers, there are two distinct windows
(Figure S6), the sizes of which are roughly
consistent across all six compounds as the window geometry is independent
of the anionic pillar: narrow, square windows where the pyridyl rings
are ortho relative to one another; larger rectangular
windows when pyridyl rings are meta relative to one
another.
Figure 2
Formula units in SIFSIX-22-Zn, TIFSIX-6-Zn, SNFSIX-2-Zn, GEFSIX-4-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Thermal ellipsoids are shown at
a probability level of 50%.
Formula units in SIFSIX-22-Zn, TIFSIX-6-Zn, SNFSIX-2-Zn, GEFSIX-4-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Thermal ellipsoids are shown at
a probability level of 50%.Pore structures were calculated from the crystallographic
data
of SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn (Poreblazer v4.0, Tables S7, S8). The
pore volume per formula unit ranges from 90.49 Å3 for TAFSEVEN-1-Zn to 135.20 Å3 for ZRFSIX-3-Zn. The maximum pore diameters vary from 4.71 to 5.92 Å (Poreblazer
v4.0), whereas the limiting pore diameters range from 3.18 to 3.72
Å.[40] The constricted pore region is
a result of the narrow pore windows occurring in the Zn2+-tepb layers while the larger pore cavity exists between
these two-dimensional layers. This degree of pore constriction is
high compared to other HCNs such as mmo topology networks
and Tripp-Cu-SIFSIX despite their unique network architectures.[23,29] The highly constricted pores present in SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn are comparable
to SIFSIX-18-M and DICRO-6-Co-i among the pcu topology HCNs which feature constricted pore environments
(Figure ). For both
of the aforementioned materials, the pore constriction arises from
the distortion of the one-dimensional metal-pillar-metal chain from
a linear to zig-zag arrangement. In SIFSIX-18-M, this
is due to the shape of the ligand, whereas in DICRO-6-Co-i it results from the nonlinear geometry of the pillar.[13,41] In contrast, the constriction of pores in SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn arises
solely from the dimensions of the Zn2+-tepb layers.
Figure 3
Accessible void surface calculated for SIFSIX-3-Zn viewed along the crystallographic (a) a- and (b) c-directions. Representations of the accessible void surface
calculated for SIFSIX-22-Zn viewed along the crystallographic
(c) a- and (d) c-directions. (e)
Plot of crystallographically determined maximum and limiting pore
diameters in representative HCN materials (red dashed line: idealized
cylindrical pores; blue dashed line: the ratio of maximum pore diameter
to pore limiting diameter in SIFSIX-22-Zn; see the SI for tabulation).
Accessible void surface calculated for SIFSIX-3-Zn viewed along the crystallographic (a) a- and (b) c-directions. Representations of the accessible void surface
calculated for SIFSIX-22-Zn viewed along the crystallographic
(c) a- and (d) c-directions. (e)
Plot of crystallographically determined maximum and limiting pore
diameters in representative HCN materials (red dashed line: idealized
cylindrical pores; blue dashed line: the ratio of maximum pore diameter
to pore limiting diameter in SIFSIX-22-Zn; see the SI for tabulation).
Characterization
Bulk samples of each compound were
synthesized by room-temperature slurry methods and characterized by
PXRD and TGA (Figures S9–S14). PXRD
patterns collected after immersion under MeOH for 1 week revealed
that all samples had retained their crystal structures. PXRD patterns
collected after exposure to accelerated stability test conditions
at 75% R.H. and 40 °C after 1 day and 1 week revealed that SIFSIX-22-Zn underwent hydrolysis and an associated phase
change within 1 day, while GEFSIX-4-Zn underwent a phase
change between 1 day and 1 week. No change was observed in the PXRD
patterns of TIFSIX-6-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn after 1 week,
indicating that these pillars provide enhanced hydrolytic stability.
Gas Sorption
After activating as-synthesized bulk samples
of each HUM, CO2 sorption isotherms were measured at 195
K to determine their textural characteristics. In all six materials,
type I isotherms with steep uptake at low pressure were observed with
saturation uptakes of 150–200 cm3 g–1 (Figure a). Brunauer–Emmett–Teller
(BET) surface areas of 387.2, 396.8, 700.0, 615.4, 625.9, and 627.5
m2 g–1 were determined for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn, respectively.
The Horvath–Kawazoe plots obtained from 195 K CO2 data revealed a narrow range of maximum pore widths between 3.7
and 4.1 Å (Figure b). These values experimentally confirm the categorization of these
materials as ultramicroporous and lie within the range of crystallographically
determined pore dimensions (Tables S7–S9). N2 sorption isotherms were also collected and afforded
BET surface area values of 414.3, 478.1, 925.1, 798.1, 1062.0, and
652.9 m2 g–1 for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn, respectively
(Figures S16–S21). The variations
between saturation CO2 and N2 uptakes are attributed
to the differing sizes and quadrupole moments of N2 and
CO2 and their interactions with the differing electrostatics
of the surfaces of each adsorbent.[42]
Figure 4
(a) 195 K CO2 sorption isotherms on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn; (b) Horvath–Kawazoe
pore-size distribution plots obtained using 195 K CO2 isotherms.
298 K sorption isotherms of (c) CO2, (d) C2H2, (e) C2H4, and (f) C2H6 on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn.
(a) 195 K CO2 sorption isotherms on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn; (b) Horvath–Kawazoe
pore-size distribution plots obtained using 195 K CO2 isotherms.
298 K sorption isotherms of (c) CO2, (d) C2H2, (e) C2H4, and (f) C2H6 on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn.Following these observations, we investigated the
room-temperature
sorption properties of each HUM toward CO2 and C2 hydrocarbons.
The CO2, C2H2, C2H4, and C2H6 sorption isotherms for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, and ZRFSIX-3-Zn are well-defined
Langmuir-type profiles (Figure c–f). PXRD data collected after sorption experiments
support the apparent reversibility of the isotherms and no loss of
crystallinity was observed (Figures S9–S14). CO2 uptakes at 298 K varied from a maximum of 99 cm3 g–1 in GEFSIX-4-Zn to a minimum
of 81 cm3 g–1 in ZRFSIX-3-Zn. Similarly, 298 K C2H2 uptakes ranged from
127 cm3 g–1 in TIFSIX-6-Zn to 102 cm3 g–1 in ZRFSIX-3-Zn. C2H4 and C2H6 isotherms
exhibited similar profiles, with lower uptakes overall vs. C2H2. The only deviation from ideal Langmuir-type
profiles was seen in the CO2 isotherm of TAFSEVEN-1-Zn, in which a minor inflection occurred at ca. 0.55 bar and 298 K.
We attribute this anomaly to the five equatorial fluorides in the
TaF72– pillar leading to a distinct electrostatic
distribution vs MF62– pillars, thereby
impacting sorption through F···HAr contacts
to the tepb ligands by altering ligand conformation and
pore dimensions.Isosteric heats of sorption (Qst) were
calculated for CO2 and C2 gases for all six adsorbents
(Figure a,b). Low
loading Qst values for CO2 varied
from 43.3 and 42.6 kJ mol–1 for SNFSIX-2-Zn and ZRFSIX-3-Zn, respectively, to 30.4 and 24.7 kJ
mol–1 for TIFSIX-6-Zn and SIFSIX-22-Zn, respectively. Low loading Qst values
for C2H2 were determined to be relatively higher
with a narrower range, from 44.8 kJ mol–1 for TIFSIX-6-Zn and GEFSIX-4-Zn to 36.5 kJ mol–1 for SIFSIX-22-Zn. In contrast, Qst values for C2H4 and
C2H6 range from 33.3 to 31.6 kJ mol–1 for C2H4 and 32.7 to 31.1 kJ mol–1 for C2H6 (Figures S22, S23). Overall, the affinity for C2H2 was
highest, encouraging us to evaluate the selectivity of the six HUMs
in the context of C2H2/CO2 and C2H2/C2H4 separations.
Figure 5
Isosteric heats
of adsorption of (a) CO2 and (b) C2H2 on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Ideal adsorbed solution theory
(IAST) selectivity determined for (c) 1:1 C2H2/CO2 (SAC) and (d) 1:99 C2H2/C2H4 (SAE) for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Comparative plots of leading HCNs
with respect to (e) SAC and (f) SAE versus C2H2 uptake.
Isosteric heats
of adsorption of (a) CO2 and (b) C2H2 on SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Ideal adsorbed solution theory
(IAST) selectivity determined for (c) 1:1 C2H2/CO2 (SAC) and (d) 1:99 C2H2/C2H4 (SAE) for SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, ZRFSIX-3-Zn, and TAFSEVEN-1-Zn. Comparative plots of leading HCNs
with respect to (e) SAC and (f) SAE versus C2H2 uptake.Calculations conducted using ideal adsorbed solution
theory (IAST)
indicate that C2H2/CO2 (1:1) selectivities
(SAC) vary between 6.5 (SIFSIX-22-Zn) and 3.8 (TAFSEVEN-1-Zn) at 1 bar. C2H2/C2H4 (1:99) selectivities (SAE) at 1 bar vary between 18.7 (SIFSIX-22-Zn) and 4.7 (TAFSEVEN-1-Zn) (Figure c,d). Perhaps most notable is that, despite
the absence of any apparent correlation between pure component Qst values for adsorbents and different gases, SAC and SAE values
follow a clear trend, that is, SIFSIX-22-Zn > SNFSIX-2-Zn > TIFSIX-6-Zn > GEFSIX-4-Zn > ZRFSIX-3-Zn > TAFSEVEN-1-Zn. This
correlation suggests that the
identity of inorganic anions is responsible for the varying affinity
toward acetylene in these HUMs and that the sorption properties are
impacted by incorporation of different pillars. No correlation between SAC and SAE values
with the electronegativity of the central atom of the anion was evident,
suggesting a need for calculation of pore surface charges and in-depth
detailed computational studies on this system to fully elucidate the
observed affinities toward C2H2.The calculated
selectivity values of these adsorbents are moderate
but when viewed together with their relatively high uptakes, it is
apparent that these fsc networks address the trade-off
between selectivity and uptake (Figure e,f). When compared to other C2H2 selective sorbents, SIFSIX-22-Zn and TIFSIX-6-Zn show a rare combination of strong selectivity and high uptake, indicating
that further exploration of this platform of materials has the potential
to produce adsorbents with strong overall performance.
Conclusions
We report herein the highly modular fsc topology HUM
platform which enabled us to explore the effect of changing inorganic
pillar on gas sorption properties. The use of SiF62–, TiF62–, GeF62–, SnF62– ZrF62–, and TaF72– in combination with Zn2+ and the tetratopic tepb ligand afforded a family of HUMs: SIFSIX-22-Zn, TIFSIX-6-Zn, GEFSIX-4-Zn, SNFSIX-2-Zn, and ZRFSIX-3-Zn and the first TaF7-pillared HUM, TAFSEVEN-1-Zn The crystal structures were determined by single-crystal
X-ray diffraction and enabled systematic studies of structure–property
relationships. Each framework features narrow pore windows, yet significant
pore cavities between the zinc(II)-tepb layers. This
is unusual in HUMs and presents a general strategy that could address
the trade-off between uptake and selectivity that is common in adsorbents.
Variation of the inorganic pillar resulted in a trend in relative
C2H2 affinity as follows: SIFSIX-22-Zn > SNFSIX-2-Zn > TIFSIX-6-Zn > GEFSIX-4-Zn > ZRFSIX-3-Zn > TAFSEVEN-1-Zn. Overall,
this work highlights the modular nature of the fsc HUM
platform and that substitution of the inorganic pillar impacts structure–property
relationships.
Authors: Stephen D Burd; Shengqian Ma; Jason A Perman; Benjamin J Sikora; Randall Q Snurr; Praveen K Thallapally; Jian Tian; Lukasz Wojtas; Michael J Zaworotko Journal: J Am Chem Soc Date: 2012-02-15 Impact factor: 15.419
Authors: Amrit Kumar; David G Madden; Matteo Lusi; Kai-Jie Chen; Emma A Daniels; Teresa Curtin; John J Perry; Michael J Zaworotko Journal: Angew Chem Int Ed Engl Date: 2015-10-06 Impact factor: 15.336
Authors: Mona H Mohamed; Sameh K Elsaidi; Lukasz Wojtas; Tony Pham; Katherine A Forrest; Brant Tudor; Brian Space; Michael J Zaworotko Journal: J Am Chem Soc Date: 2012-11-21 Impact factor: 15.419
Authors: Soumya Mukherjee; Nivedita Sikdar; Daniel O'Nolan; Douglas M Franz; Victoria Gascón; Amrit Kumar; Naveen Kumar; Hayley S Scott; David G Madden; Paul E Kruger; Brian Space; Michael J Zaworotko Journal: Sci Adv Date: 2019-11-29 Impact factor: 14.136
Authors: Hayley S Scott; Naoki Ogiwara; Kai-Jie Chen; David G Madden; Tony Pham; Katherine Forrest; Brian Space; Satoshi Horike; John J Perry Iv; Susumu Kitagawa; Michael J Zaworotko Journal: Chem Sci Date: 2016-05-10 Impact factor: 9.825